Metadata-Version: 2.4
Name: symbolmachines
Version: 0.1.0
Summary: Symbol Machines Memory SDK — Long-term cognitive memory for any LLM.
Author-email: Symbol Machines <joe@symbolmachines.com>
License: MIT
Project-URL: Homepage, https://symbolmachines.com
Project-URL: Repository, https://github.com/cojohen/symbol-sdk/python
Keywords: LLM,memory,knowledge-graph,symbol-machines,AI,AGI
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.8
Description-Content-Type: text/markdown
Requires-Dist: requests>=2.25

# Symbol Machines Memory SDK (Python)

A lightweight SDK that adds structured long-term memory to any LLM.

Learn more at https://symbolmachines.com

---

## Installation

```bash
pip install symbolmachines
```

## Usage

```python
from symbolmachines import Symbol

symbol = Symbol(api_key="sk-...")

# Inject memory context before calling your LLM
context = symbol.memory.inject.create(input="What should i eat for lunch?")

# ... call your LLM with the contextual input ...

# Save memory after generation
symbol.memory.save.create(output="Eating sushi is a healthy meal.")
```

---

## API

`symbol.memory.inject.create(input=str)`
Injects memory context into user input.

`symbol.memory.save.create(output=str)`
Writes an output into the Symbol's memory.

---

## Configuration

```python
symbol = Symbol(
    api_key="sk-...",
    base_url="https://api.symbol.ai/v1" # optional override
)
```
